Research ArticleIL-37 Exerts Anti-Inflammatory Effects in Fetal Membranes ofSpontaneous Preterm Birth via the NF-κB and IL-6/STAT3Signaling Pathway

Lulu Wang,1,2,3 Zheng Liu,1,2,3 Dongni Huang,1,2,3 Yuxin Ran,1,2,3 Hanwen Zhang,1,2,3

Jie He,1,2,3 Nanlin Yin ,1,2,3,4 and Hongbo Qi 1,2,3

1Department of Obstetrics, The First Affiliated Hospital of Chongqing Medical University, Chongqing 400016, China2Chongqing Key Laboratory of Maternal and Fetal Medicine, Chongqing Medical University, Chongqing 400016, China3Joint International Research Laboratory of Reproduction and Development of Chinese Ministry of Education,Chongqing Medical University, Chongqing 400016, China4Center for Reproductive Medicine, The First Affiliated Hospital of Chongqing Medical University, Chongqing 400016, China

Correspondence should be addressed to Nanlin Yin; yinnanlin@cqmu.edu.cn and Hongbo Qi; qihongbocy@gmail.com

Received 24 February 2020; Revised 4 June 2020; Accepted 8 June 2020; Published 11 July 2020

Academic Editor: Markus H. Gräler

Copyright © 2020 Lulu Wang et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Spontaneous preterm birth (sPTB), defined as delivery before 37 weeks of gestation, is thought to be a multifactorial syndrome.However, the inflammatory imbalance at the maternal-fetal interface promotes excessive secretion of inflammatory factors andinduces apoptosis and degradation of the extracellular matrix (ECM), which can subsequently lead to preterm birth. As an anti-inflammatory molecule in the IL-1 family, interleukin-37 (IL-37) mainly plays an inhibiting role in a variety of inflammatorydiseases. However, as a typical inflammatory disease, no previous studies have been carried out to explore the role of IL-37 insPTB. In this study, a series of molecular biological experiments were performed in clinical samples and human amnioticepithelial cell line (Wistar Institute Susan Hayflick (WISH)) to investigate the deficiency role of IL-37 and the potentialmechanism. Firstly, the results indicated that the expression of IL-37 in human peripheral plasma and fetal membranes wassignificantly decreased in the sPTB group. Afterward, it is proved that IL-37 could significantly suppress the production oftumor necrosis factor-α (TNF-α), interleukin-1β (IL-1β), and interleukin-6 (IL-6) in WISH cells. Simultaneously, once silenceIL-37, LPS-induced apoptosis and activity of matrix metalloproteinases (MMPs) 2 and 9 were significantly increased. Inaddition, the western blot data showed that IL-37 performed its biological effects by inhibiting the NF-κB and IL-6/STAT3pathway. In conclusion, our results suggest that IL-37 limits excessive inflammation and subsequently inhibits ECM remodelingand apoptosis through the NF-κB and IL-6/STAT3 signaling pathway in the fetal membranes.

1. Introduction

Spontaneous preterm birth (sPTB), occurring before 37weeks of gestation, is a specific disorder of pregnancy. Witharound 15 million preterm infants born each year, account-ing for approximately 10.6% of all live births worldwide,sPTB is considered to be the most frequent cause of mortalityin neonatal and the second in children age below 5 years oldworldwide [1, 2]. Furthermore, those survived preterm neo-nates have higher rates of multiple complications, includingneurological dysplasia, respiratory distress syndrome, neona-

tal pneumonia, necrotizing enterocolitis, and retinal disease,which cause a heavy burden to the family and society [3, 4].sPTB is considered to be a multifactorial syndrome, relatedto aging, functional progesterone withdrawal, cervical dis-ease, and stress [5]. However, its etiology and pathogenesisremain unclear, which attracts more attention in the latestyears.

Many studies have mentioned that the pathogenesis ofpreterm birth is closely related to changes in the microenvi-ronment of the maternal-fetal interface. Gomez-Lopez et al.and Boyle et al. have pointed out that it is prone to form an

HindawiMediators of InflammationVolume 2020, Article ID 1069563, 15 pageshttps://doi.org/10.1155/2020/1069563

inflammation cascade, which leads to the occurrence of pre-term birth when the inflammatory balance is disrupted at thematernal-fetal interface [6, 7]. We have known that thematernal-fetal interface is mainly composed of fetal mem-brane, decidual, and uterine. Among them, the fetal mem-brane is unique since it is not only functioning as anessential component separating the fetus from the mother’sbody, but also operating as a vital window for understandingthe pathophysiology of preterm birth [8]. The fetal mem-brane is rich in the extracellular matrix (ECM). The degrada-tion of ECM is often closely related to the incidence ofpreterm birth [9]. Furthermore, among all the different cellsof the fetal membrane, amniotic epithelial cells have attractedthe most attention since it is considered to be the earliest cellto receive fetal-derived signals [10]. The apoptosis of humanamniotic epithelial cells is involved in preterm birth aswell [11]. Meanwhile, human amniotic epithelial cells canbe easily activated by external stimuli including lipopoly-saccharide (LPS), hormone, and mechanical traction, andthen subsequently releasing a large number of inflamma-tory factors, such as tumor necrosis factor-α (TNF-α),interleukin-1β (IL-1β), and interleukin-6 (IL-6), whichultimately mediate the excessive inflammation associatedwith preterm birth [12, 13].

Therefore, it is obvious to speculate that the most key linkin excessive inflammation is the release of a large number ofproinflammatory factors. In recent years, with the accumula-tion of data, it is found that interleukins play a critical role inregulating the release of proinflammatory factors. Forinstance, IL-33 could induce the expression ofprogesterone-induced blocking factor (PIBF1) by function-ing the prevention effect of decidual B cells from pretermlabor caused by LPS-induced proinflammatory factors [14].Dambaeva et al. found that IL-22 can also prevent pretermbirth and intrauterine fetal death by reducing the productionof proinflammatory factors exposed to LPS [15]. As a typicalrepresentative of proinflammatory factors, IL-1 and its fam-ily members orchestrate numbers of inflammatory diseases,including preterm birth. However, interleukin-37 (IL-37) isa unique member of the IL-1 family, discovered in silico in2000, which functions as a natural suppressor of inflamma-tory and immune responses. The gene of IL-37 is located atchromosome 2 and the molecular weight is about17∼25 kDa.The structure of IL-37 is similar to the IL-1 fam-ily, which consists of twelve β-tube-like structures. It has sixexons encoding five subtypes of IL-37, which are named asIL-37a, IL-37b, IL-37c, IL-37d, and IL-37e. In addition, IL-37 can be found in multiple human tissues, such as lung,thymus, bone marrow, and uterus tissues [16, 17]. Underphysiological conditions, the expression of IL-37 is at a lowlevel. However, once stimulated by inflammation, theexpression of IL-37 will increase sharply [18]. Ellisdon et al.mentioned that IL-37 could significantly ameliorate LPS-induced inflammation in sepsis [19]. And Ye et al. indicatedthat IL-37 could inhibit the production of inflammatorycytokines in systemic lupus erythematosus [20]. Moreover,in the perinatal field, Southcombe and Yu et al. have dem-onstrated that IL-37 could be aberrantly expressed on theplacenta and umbilical cord tissues in patients with pre-

eclampsia and gestational diabetes mellitus, respectively[21, 22]. These studies mainly focus on the differences inexpression. Little is known about the mechanisms. How-ever, preterm birth is considered as a typical inflammatorydisease, and no data have been found to reveal the role ofIL-37. Therefore, we infer whether a deficiency of IL-37 inhuman fetal membranes is related to the pathogenesis ofsPTB. This study is aimed at demonstrating that IL-37can limit excessive inflammation and subsequently inhibitinflammation-induced ECM remodeling and apoptosis ofhuman amniotic epithelial cells through the NF-κB andIL-6/STAT3 pathway, which might further enrich theoret-ical strategies for sPTB.

2. Materials and Methods

2.1. Sample Collection. Subjects were enrolled from the FirstAffiliated Hospital of Chongqing Medical University andethics approval was gained from the Ethics Committee ofthe First Affiliated Hospital of Chongqing Medical Univer-sity. All the participants signed informed consent andrecruited into the study if they were a singleton pregnancy.sPTB was defined according to the guidelines published bythe American Journal of Obstetrics and Gynecology. All par-ticipants with chronic medical disorders were excluded,including diabetes mellitus, cardiovascular disease, infectiousdisease, autoimmune disease, chronic renal disease, chronichypertension, and metabolic diseases. Fetal membrane tis-sues were collected immediately after delivery, and bloodsamples were taken from the vein from both patients withsPTB (n = 19) and control pregnant women (n = 21). Clinicalinformation of study objects is shown in Table 1. Afterobtaining the samples, they were transferred to the labora-tory with ice boxes within 1 hour and washed in sterile phos-phate buffer solution for three times in order to remove thebloodstain on tissues. Blood samples were centrifuged at3000 rpm for 10 minutes at 4°C. All the samples were storedat -80°C before analysis.

2.2. Cell Line and Cell Culture. The WISH (Wistar InstituteSusan Hayflick) cell line was purchased from the commer-cialized company. Cells were cultured in MEM (Gibco,USA) complete medium containing 10% fetal bovineserum (PAN, Germany), 1% nonessential amino acid(Solarbio Biotechnology, Beijing, China), and 1% penicillin(110U/ml) and streptomycin (100μg/ml) (BeyotimeBiotechnology, Shanghai, China). Then, cells were incu-bated in 5% CO2 at 37°C.

2.3. Western Blot. Human amniotic epithelial cells treatedwith recombinant IL-37 protein (rhIL-37, R&D Systems,USA), recombinant IL-6 protein (rhIL-6, PeproTech, USA),and fetal membrane tissues were lysed in RIPA lysis buffer(MedChemExpress, USA). Total protein concentration wasdetected utilizing a BCA Protein Assay Kit (BeyotimeBiotechnology, Shanghai, China). The antibodies againstthe following proteins were used: IL-37 (1 : 500; Santa Cruz,USA); Bax, Bcl-2, MMP9, T-NF-κB, p-NF-κB (Ser536), β-actin (1 : 1000; Cell Signaling Technology, USA); MMP2,

2 Mediators of Inflammation

STAT3, p-STAT3 (Tyr705) (1 : 1000; Affinity Biosciences,USA); β-tubulin, GAPDH (1 : 1000; Servicebio Biotechnology,Wuhan, China). The same volume of protein (40μg) was sep-arated electrophoretically by SDS-PAGE and transferred to aPVDF membrane (Roche, Germany). After the transfer, themembranes were blocked with 5% skimmed milk for 1 hourand incubated overnight at 4°C with the indicated antibody.The next day, the membranes were incubated with an appro-priate horseradish peroxidase-linked secondary antibody for 1hour at room temperature. After incubation with the second-ary antibody, the bands were visualized by an ECL. Thesecondary antibodies utilized in this research were goat anti-rabbit IgG-HRP (1 : 5000; Boster, Beijing, China) and goatanti-mouse IgG-HRP (1 : 5000; ZSGB-BIO, Beijing, China).Image Lab Software (Bio-Rad, California, USA) was used forimaging and ImageJ was used for densitometric analysis.

2.4. Quantitative RT-PCR. Total RNA was extracted fromfetal membrane tissues and cells using TRIzol reagent(Thermo Fisher Scientific, USA) and quantified by measur-ing the ratio of A260nm/A280nm using NanoDrop (ThermoFisher Scientific, USA). Total RNA was reverse transcribedusing an EvoScript Universal cDNA Master Reagent Kit(Roche, Germany). The primers used in this study are listedin Table 2 (TSINGKE, Chongqing, China). Briefly, the real-time PCR reaction was consisted of 2μl of cDNA, 1μl of eachprimer, 5μl of SYBR Green qPCR Master Mix (MedChem-Express, USA), and 1μl of ultrapure water, for a total volumeof 10μl, and was conducted in a Bio-Rad CFX-96 system(Bio-Rad, California, USA). The reaction conditions were asfollows: 5 minutes at 95°C, 40 cycles of 10 seconds at 95°C,20 seconds at 63.3°C, and 20 seconds at 72°C, respectively;melting curves were generated after each endpoint amplifi-cation for 10 seconds at 95°C, followed by 30-second incre-ments of 0.5°C from 65 to 95°C. β-Actin was used as anendogenous reference. The 2−△△Ct method was used forthe calculation.

2.5. Immunofluorescence. When the bloodstain was removedfrom the tissue, the fetal membrane was fixed with 4% para-formaldehyde for 24 hours, and then dehydrated in 30%sucrose solution for 6 hours. Lastly, the tissues were embed-ded in optimal cutting temperature compound and slicedinto sections. When conducting immunofluorescence, theslices firstly permeabilized with 0.1% Triton X-100 for 5minutes and then blocked with 5% BSA for 1 hour at 37°C.The tissues were incubated overnight at 4°C with target anti-bodies at an appropriate dilution. The target antibodies IL-37(1 : 50; Santa Cruz, USA) were used. The next day, tissueswere dealt with the appropriate fluorescein-conjugatedsecondary antibodies against rabbit or mouse IgG (1 : 100;Abbkine Biotechnology, Wuhan, China). 6-Diamidino-2-phenylindole (DAPI, Servicebio Biotechnology, Wuhan,China) was utilized for nuclear staining.

2.6. Immunohistochemistry. Fetal membranes were fixed with4% paraformaldehyde, embedded in paraffin, and sliced at athickness of 3-5μm. The slices were dewaxed in xylene anddehydrated in gradient alcohol solutions. As to antigenretrieval, the slices were heated in sodium citrate buffer athigh temperature for 4 minutes and middle temperature for15 minutes in a microwave oven. Then, the endogenous per-oxidase activity was removed with 3% H2O2 for 10 minutesand blocked with goat serum for 1 hour. Slices were incu-bated with IL-37 antibody (1 : 50, Santa Cruz, USA) at 4°Covernight. The next day, the sections were incubated with abiotin-labeled secondary antibody (ZSGB-BIO, Beijing,China). 3,3′-Diaminobenzidine (DAB) and hematoxylin(blue) were utilized to label the target and cell nuclei, respec-tively. Lastly, the stained sections were then evaluated by aSuper Sensitive™ Link-Label IHC detection System(BioGenex, San Ramon, CA, USA).

2.7. Cell Apoptosis Assay. According to the manufacturer’sprotocols of the Annexin V-FITC kit (Beyotime Biotechnol-ogy, Beijing, China), the cells were plated into 6-well plates at4 × 105 cells/ml per well. Apoptosis was assessed by flowcytometry after the cells were transfected with IL-37 siRNAfor 48 hours. The cells were collected and washed three timeswith cold phosphate buffer solution, and then mixed with theAnnexin V-FITC and PI binding buffer for 20 minutes.Finally, the mixture was analyzed using a flow cytometer(BD Biosciences, San Jose, CA, USA).

2.8. EdU DNA Synthesis Assay. After transfected with IL-37siRNA for 48 hours, cell proliferation assay was performedusing a 5-ethynyl-20-deoxyuridine (EdU) detection kit(Ribobio, Guangzhou, China). According to the manufac-turer’s instructions, cells were dealt with 50μmol/l EdU for2 hours at 37°C and then fixed with 4% paraformaldehydefor 30 minutes. After, cells were incubated with 2mg/ml gly-cine for 5 minutes before treated with 0.5% Triton X-100 inphosphate buffer solution for 10 minutes and stained with1× Apollo reaction cocktail for 30 minutes at room tempera-ture. Subsequently, the counterstaining of cell nuclei wasconducted with Hoechst. The proliferation rate was

Table 1: Clinical information of study objects.

Maternalcharacteristics

Normal pregnancy(n = 21)

Spontaneous pretermbirth (n = 19)

Maternal age(years)

27:86 ± 2:67 28:16 ± 3:18

Gestational age(weeks)

39:78 ± 0:78 33:40 ± 3:30b

Prepregnancy BMI(kg/m2)

20:80 ± 2:30 21:17 ± 2:36

Neonatal birthweight (g)

3351 ± 395:80 2337 ± 525:90b

fFN (positive ornegative)

Negative Negative

Placentalweight (g)

586:20 ± 104:30 496:30 ± 85:26a

BMI: body mass index. Data are presented as the mean ± SD. ap < 0:01;bp < 0:0001.

3Mediators of Inflammation

determined by the ratio of EdU staining-positive cells to totalcells. Images were captured under a fluorescence microscope.

2.9. Cell Viability Assay. WISH cells were planted into 96-well plates (5000 cells/well) and then transfected with IL-37siRNA for 48 hours after adhesion. Two fluorescent probes(Live/Death Cell Imaging Kit, Invitrogen, USA) (1drop/500μl medium) were used in each well and then incu-bated for 15 minutes prior to photography under fluores-cence microscopy. Six fields of view were captured for eachgroup. Then, the counts of total and dead cells werecalculated.

2.10. Transfection of siRNA. The sequences employed wereIL-37 siRNA, 5′-UCUACUGUGACAAGGAUAATT-3′and negative control siRNA (si_NC), 5′-UUCUCCGAACGUGUCACGUTT-3′. Those were designed and synthesizedby GenePharma (Shanghai, China). Cells were planted into6-well plates at a density of 105 cells/well, and then trans-fected with IL-37 siRNA or negative control siRNA usingLipo 2000 reagent according to the recommended protocol.After transfection for 24 hours, cells were prepared for RT-qPCR and 48 hours for western blot. The transfection effi-ciency was confirmed by RT-qPCR and western blot.

2.11. ELISA Assay. The concentration of IL-37 in humanperipheral blood plasma was measured by a human IL-37ELISA kit (RayBiotech, USA). The ELISA kit used to measurethe expression of IL-6 in cell supernatant was purchasedfrom Elabscience (Wuhan, China), and the ELISA kits usedto measure the expression of IL-1β, TNF-α in cell superna-tant were purchased from 4A Biotech (Beijing, China). Allthe experiments were conducted according to the manufac-turer’s instructions. The absorbance of the supernatant wasassessed at 450nm in a MultiskanGO plate reader (ThermoFisher Scientific, Waltham, MA, USA).

2.12. Gelatin Zymography. The total proteins of fetal mem-branes and cells were extracted using the same method asthe western blot. The proteins were separated by 10% poly-acrylamide gels containing sodium dodecyl sulfate (SDS)and 2μg/ml gelatin. After electrophoresis, the gels werewashed with 2.5% Triton X-100 for 1.5 hours to remove theSDS and incubated for 48 hours with an incubating solutioncontaining 50mM Tri-HCl, 200mMNaCl, 5mM CaCl2, and0.02% Brij-35 at 37°C. After incubation, the gels were stained

in 0.5%w/v Coomassie brilliant blue R-250 at 37°C for 1 hourand then incubated with a decolorizing solution to visualizegelatinolytic bands. The different protease bands were quali-tatively and quantitatively analyzed with Image Gel software(Bio-Rad, California, USA).

2.13. Statistical Analysis. Student’s t-test was applied to deter-mine the significant differences between the two groups.Comparison among multiple groups was performed byone-way ANOVA to detect the significant differences. Alldata were shown as the mean ± standard deviation (SD). Allexperiments were repeated at least three times. All statisticalanalysis was performed utilizing GraphPad Prism software,version 8.0 (GraphPad Software Inc., La Jolla, CA, USA).Results were considered statistically significant whenp < 0:05 level.

3. Results

3.1. IL-37 Expression Significantly Decreases in sPTBPregnancies. Firstly, we compared the expression of IL-37in the preterm and the term groups. The results showed thatIL-37 was significantly decreased in the peripheral bloodplasma of sPTB compared to term labor [0:89 ± 0:18 vs.0:52 ± 0:04, t = 2:094, p = 0:046; (Figure 1(a))]. In addition,the mRNA and protein expression levels of IL-37 were alsolower in fetal membranes of the sPTB group than those ofthe term group [2:46 ± 0:38 vs. 0:61 ± 0:16, t = 3:520,p = 0:0022; Figure 1(b); 0:92 ± 0:10 vs. 0:39 ± 0:06, t = 4:419, p = 0:0013; Figures 1(c) and 1(d)]. Moreover, immunofluo-rescence (IF) and immunohistochemistry (IHC) were uti-lized. As shown in Figures 1(e) and 1(f), IL-37 waslocalized at human amnion and chorion layers. Collectively,these results indicated that the expression of IL-37 in humanperipheral plasma and fetal membranes was significantlylower in the sPTB group than that in the term group.

3.2. IL-37 Limits Excessive Inflammation. After we deter-mined that IL-37 did significantly decrease in fetal mem-branes of the sPTB group, we are wondering how it works.To determine the effect of IL-37 on cytokine production inWISH cells, quantitative real-time PCR and ELISA were usedto screen various inflammatory cytokines. When WISH cellswere treated with both LPS (1μg/ml) and rhIL-37 (10 ng/ml,50 ng/ml, 100 ng/ml) for 6 hours, 12 hours, and 24 hours, themRNA and protein expressions of IL-1β, IL-6, and TNF-α

Table 2: Characteristic of primers.

Genes Sense primer (5′→ 3′) Antisense primer (5′→ 3′)IL-1β CCACAGACCTTCCAGGAGAAT GTGCACATAAGCCTCGTTATCC

IL-6 CCTAGAGTACCTCCAGAACAGA CAGGAACTGGATCAGGACTTT

TNF-α ACCTCTCTCTAATCAGCCCTCT GGGTTTGCTACAACATGGGCTA

IL-37 TTCTTTGCATTAGCCTCATCCTT CGTGCTGATTCCTTTTGGGC

MMP2 AAGGACAGCCCTGCAAGTTT GTTCCCACCAACAGTGGACA

MMP9 GGTGATTGACGACGCCTTTG GGACCACAACTCGTCATCGT

β-Actin TGGCACCCAGCACAATGAA CTAAGTCATAGTCCGCCTAGAAGCA

4 Mediators of Inflammation

IL-3

7 co

ncen

trat

ion

(ng/

ml)

0.50.0

Term sPTB

1.0

1.5

2.0

2.5 ⁎

(a)

Term sPTB

IL-3

7 m

RNA

expr

essio

n(a

rbitr

ary

unity

)

2

4

6

0

⁎⁎

(b)

sPTBTerm

𝛽-Actin

IL-37

(c)

0.0

0.5

1.5

IL-3

7 to

𝛽-a

ctin

ratio

n(a

rbitr

ary

unity

)

1.0

Term sPTB

⁎⁎

(d)

200 𝜇m 200 𝜇m 200 𝜇m

200 𝜇m200 𝜇m200 𝜇m

DAPI IL-37

sPTB

Term

Merge

(e)

100 𝜇m200 𝜇m

100 𝜇m200 𝜇m

sPTB

Term

Chorion

Amnion

Amnion

Chorion

Amniotic epithelium

Amniotic epithelium

(f)

Figure 1: The expression and distribution of IL-37 in fetal membranes and plasma complicated with spontaneous preterm birth. (a) Theplasmas from term (n = 14) and sPTB delivery women (n = 15) were collected, and IL-37 level was detected by ELISA. The detection rangeof IL-37 was 80-20000 pg/ml. (b) The mRNA expression level of IL-37 in fetal membranes was measured by quantitative real-time PCRanalysis. The results were normalized to β-actin. (c) The fetal membranes from term (n = 6) and sPTB delivery women (n = 6) wereharvested, and IL-37 level was detected by western blot. (d) Statistical analysis of western blot in the result (c). (e) Immunofluorescencestaining was utilized to locate the expression site of IL-37 (green) in fetal membranes. Scale bar: 200μm. (f) Immunohistochemicalstaining of IL-37 in fetal membranes. Scale bar: 200 μm and 100μm. ∗p < 0:05; ∗∗p < 0:01.

5Mediators of Inflammation

decreased in a dose- and time-dependent manner(Figures 2(a)–2(d)). The most significant decrease occurredat a concentration of 100ng/ml and time point 24 hours, sothis concentration and time point were selected for furtherstudies. Next, we transfected IL-37 siRNA (si_IL-37) andnegative control siRNA (si_NC) into WISH cells, and theirtransfection efficiencies were confirmed by quantitativereal-time PCR and western blot (Figure 2(e)). Then, experi-ments were carried out. The results showed that knockdownof IL-37 significantly increased the mRNA and proteinexpressions of downstream inflammatory factors (IL-1β,IL-6, and TNF-α) when exposed to LPS, which were reversedby the administration of rhIL-37 (Figure 2(f)). However, theIL-37 siRNA group no longer downregulated the expressionof IL-1β, IL-6, and TNF-α in the absence of LPS stimulation.Dramatically, among the three inflammatory factors, thechange of IL-6 was the most obvious one.

3.3. IL-37 Inhibits ECMDegradation. The results showed thatIL-37 can inhibit the release of downstream inflammatoryfactors in fetal membranes and thus limit excessive inflam-mation. Previous studies have revealed that IL-37 can inhibitthe release of matrix metalloproteinases (MMPs) in otherdiseases, such as endometriosis [23]. Therefore, we specu-lated that IL-37 will regulate the effect of MMPs in pretermbirth. Firstly, the role of MMP2 and MMP9 was verified(Figure 3(a)). As presented in Figure 3(b), the activity ofMMP9 was notably elevated in fetal membranes of the pre-term birth group. Then, the effect of rhIL-37 on MMP2 andMMP9 activity and expression inWISH cells were measured.Our data indicated that rhIL-37 treatment significantlyreduced the mRNA expressions of MMP2 and MMP9induced by LPS (Figure 3(d)). However, similar results wereonly observed in the protein level of MMP9 (Figures 3(e)and 3(f)). AlthoughMMP2 was also decreased, the differencewas not statistically significant. In addition, rhIL-37 signifi-cantly reduced the LPS-induced activity of MMP9, whereasMMP2 was not affected (Figure 3(c)). Correspondingly, thesiRNA-mediated knockdown of IL-37 significantly upregu-lated the activity of MMP9, whereas had no effect onMMP2 (Figures 3(g) and 3(h)), which was in accordancewith the results found in mRNA (Figure 3(i)) and proteinlevels (Figures 3(j) and 3(k)). However, the administrationof rhIL-37 alone produced no effect on the expression andactivity of MMP2 and MMP9 (Figures 3(c)–3(f)).

3.4. IL-37 Inhibits Apoptosis of Human Amniotic EpithelialCells. Cell proliferation and apoptosis maintain life activities.However, the onset of preterm birth is inseparable from theproliferation and apoptosis of amniotic epithelial cells [9].Therefore, the effect of IL-37 on the proliferation and apo-ptosis of amniotic epithelial cells was investigated. EdU stain-ing showed that knockdown of IL-37 significantly reducedDNA synthesis in WISH cells [51:43% ± 0:52 vs. 32:54% ±0:34, p < 0:0001; Figures 4(a) and 4(b)]. Furthermore, theratio of dead cells to total cells was observably increased inthe IL-37 siRNA group compared with the si_NC group[1:71% ± 0:21 vs. 4:69% ± 0:84, p < 0:0001; Figures 4(c) and4(d)]. Hence, it appears that IL-37 deficiency can not only

inhibit proliferation but also promote cell death. In line withthese data, the percentage of apoptotic cells was detectedremarkably augmented with flow cytometry after the trans-fection of IL-37 siRNA into WISH cells [5:11% ± 0:61 vs.10:70% ± 3:26, p = 0:0276; Figures 4(e) and 4(f)]. In orderto identify the influence of IL-37 on apoptosis-related pro-teins, the expression of Bcl-2 and Bax was analyzed by west-ern blot (Figure 4(g)). As shown in Figure 4(h), theexpression of Bax was significantly upregulated in the IL-37siRNA group when contrast to the si_NC group, while Bcl-2 decreased. Additionally, compared with the si_NC group,phosphorylated STAT3 (p-STAT3) was activated in the IL-37 siRNA group (Figures 4(i) and 4(j)). These data indicatedthat IL-37 could efficiently inhibit apoptosis by regulatingantiapoptotic molecule Bcl-2 and proapoptotic moleculeBax via the STAT3 pathway.

3.5. IL-37 Exerts Anti-Inflammatory Effects via the NF-κB andIL-6/STAT3 Signaling Pathway. To further explore themolecular mechanism of IL-37 on the process of inflamma-tion induced by LPS, the effects of IL-37 on the NF-κB andSTAT3 signaling in WISH cells were examined(Figure 5(a)). The western blot data showed that rhIL-37obviously attenuated phosphorylation of the NF-κB andSTAT3 induced by LPS in WISH cells (Figure 5(b)). Mean-while, the WISH cells were administrated by rhIL-6 andrhIL-6+rhIL-37 to determine whether the IL-6/STAT3 sig-naling pathway was involved in the IL-37-mediated anti-inflammatory effects. As shown in Figures 5(c) and 5(d),phosphorylation of STAT3 was activated after treated withrhIL-6, while reversed by rhIL-37. These results indicatedthat IL-37 might exert its anti-inflammatory effects mainlythrough the NF-κB and IL-6/STAT3 signaling pathway inhuman amniotic epithelial cells.

4. Discussion

sPTB is a leading cause of neonatal mortality and morbidityfor children under 5 years old worldwide, which carries aheavy economic burden on families and society [24]. It is amultifactorial syndrome, and researches on the mechanismof inflammation have received widespread attention in thelatest years. The inflammatory factors such as IL-1β, TNF-α, and IL-6 were all reported to involve in the pathogenesisof preterm birth [25, 26]. Elevated concentration of TNF-αin the amniotic fluid activated nuclear factor κB (NF-κB)pathway, triggering amplification of inflammatory processesthat lead to preterm birth [27, 28]. Recently, increased IL-6levels in amniotic fluid from the second trimester were asso-ciated with the onset of sPTB, which makes it a potentialbiomarker [29, 30]. As we know, IL-1β, as the most inten-sively researched molecule in the field of preterm birth, hasalways been considered as the molecule that dominates theinflammatory cascade at the maternal-fetal interface [31].The IL-1 family is made up of 11 different cytokines, whichpresent the same structure and share a similar function.However, the pathways through which they are activatedare quite different from one to another. To our surprise, themembers of the IL-1 family can directly interact with each

6 Mediators of Inflammation

LPS+rhIL-37 10 ng

LPS+rhIL-37 100 ngLPS+rhIL-37 50 ng

0

1

2

3

4

ControlLPS

####

####

####

IL-1𝛽 IL-6 TNF-𝛼Rela

tive m

RNA

expr

essio

n(a

rbitr

ary

unit)

⁎⁎⁎

⁎⁎⁎⁎ ⁎⁎⁎⁎

⁎⁎⁎⁎

⁎⁎⁎⁎

⁎⁎⁎

(a)

LPS+

rhIL

-37

100

ng

LPS+

rhIL

-37

50 n

g

LPS+

rhIL

-37

10 n

g

LPS

Cont

rol

LPS+

rhIL

-37

100

ng

LPS+

rhIL

-37

50 n

g

LPS+

rhIL

-37

10 n

g

LPS

Cont

rol

LPS+

rhIL

-37

100

ng

LPS+

rhIL

-37

50 n

g

LPS+

rhIL

-37

10 n

g

LPS

Cont

rol

IL-6

conc

entr

atio

n(n

g/m

l)

0.0

0.5

1.0

2.5

2.0

1.5

####150

50

0

100

IL-1𝛽

conc

entr

atio

n(p

g/m

l)

##

##25

20

15

10

5

0TNF-𝛼

conc

entr

atio

n(p

g/m

l)⁎⁎⁎⁎⁎ ⁎

⁎

⁎⁎⁎⁎

(b)

2.0

0.0

0.5

1.0

1.5

IL-1𝛽

mRN

A ex

pres

sion

(arb

itrar

y un

ity)

0 h 6 h 12 h 24 h

LPSLPS+rhIL-37

IL-6

mRN

A ex

pres

sion

(arb

itrar

y un

ity)

2.0

1.5

1.0

0.5

0.0 0.0

1.5

TNF-𝛼

mRN

A ex

pres

sion

(arb

itrar

y un

it)

0.5

1.0

0 h 6 h 12 h 24 h 0 h 6 h 12 h 24 h

⁎⁎⁎⁎⁎⁎

⁎⁎⁎

⁎⁎

⁎⁎ ⁎⁎

(c)

LPSLPS+rhIL-37

60

70

80

100

IL-1𝛽

conc

entr

atio

n(p

g/m

l)

90

0 h 6 h 12 h 24 h

22

20

18

16

14

12TNF-𝛼

conc

entr

atio

n(p

g/m

l)

IL-6

conc

entr

atio

n(n

g/m

l)

0.5

1.0

2.0

1.5

0 h 6 h 12 h 24 h 0 h 6 h 12 h 24 h0.0

2.5

⁎⁎⁎

⁎⁎⁎⁎

⁎⁎⁎

⁎

(d)

Figure 2: Continued.

7Mediators of Inflammation

other. Gallenga et al. and Franza et al. have indicated that IL-37 could inhibit IL-1-mediating activities and IL-38 couldinhibit the proinflammatory activity of IL-36 [32, 33]. As amember of the IL-1 family, IL-37 exerts a completely differ-ent anti-inflammatory effect from IL-1β. Nold et al. foundthat IL-37 releasing from epithelium and macrophagesalmost completely suppressed the expression of several pro-inflammatory factors (mainly IL-1β, IL-6, and TNF-α)in vitro, and protected mice against sepsis induced by LPSin vivo [34]. It has been reported that IL-37 performs a pro-tective effect on colitis, arthritis, pancreatitis, and otherinflammatory diseases which were induced by exogenousstimuli [16]. However, the role of IL-37 in preterm birth, alsoan inflammatory disease, has not been clearly reported yet.Therefore, we hypothesized that IL-37 will affect the occur-rence of preterm birth by limiting the excessive inflammatoryeffect at the maternal-fetal interface.

Firstly, we are wondering whether IL-37 has differentexpressions in maternal tissues between the preterm birthgroup and the control group. The results showed that theexpression of IL-37 in the plasma and fetal membranes wasdecreased when compared with that in the term group, whichsuggested that IL-37 might be associated with the occurrenceof preterm birth. Next, we verify whether IL-37 really exerts

anti-inflammatory effects at the cellular level with the humanamniotic epithelial cell line (WISH). Our data showed thatIL-37 indeed suppressed the expression of inflammatory fac-tors (IL-1β, IL-6, and TNF-α). As we know, the inflamma-tory cascade, mediating by different inflammatory factors,such as IL-1β, IL-6, and TNF-α, plays a crucial role at thematernal-fetal interface, which involves in uterine contractil-ity and membrane rupture related to sPTB. Notably, amongthe three above inflammatory factors, the level of IL-6showed the most significant change in WISH cells. Previousstudies have shown that the resident cells of the chorionand decidua are capable of synthesizing and secreting IL-6,and its production is stimulated by LPS or proinflammatoryfactors such as IL-1β and TNF-α [35]. Furthermore, anotherstudy has demonstrated that IL-6 secretion is upregulated infetal membranes of preterm pregnancies, indicating that IL-6might be a key factor in accelerating the events of pretermbirth [36]. Therefore, it is understandable that the changeof IL-6 is the most obvious one among these markers. Insummary, IL-37 is capable of inhibiting the production ofIL-1β, IL-6, and TNF-α, which may play a vital role in thefield of premature treatment.

The fetal membrane is not only an important mechanicalbarrier between the mother and the fetus but also an essential

si-N

C

si-IL

-37

IL-3

7 m

RNA

expr

essio

n(a

rbitr

ary

unit)

0

4

3

2

1 IL-37𝛽-Actin

si_NC si_IL-37⁎⁎⁎⁎

(e)

150

50

0

100

4

2

1

0

30

20

10

0TNF-𝛼

conc

entr

atio

n(p

g/m

l)

####ns

####

ns

####ns3

IL-6

conc

entr

atio

n(n

g/m

l)

LPS+

rhIL

-37

LPS+

si_IL

-37

LPS

si_N

Csi_

IL-3

7

Cont

rol

LPS+

rhIL

-37

LPS+

si_IL

-37

LPS

si_N

Csi_

IL-3

7

Cont

rol

LPS+

rhIL

-37

LPS+

si_IL

-37

LPS

si_N

Csi_

IL-3

7

Cont

rolIL

-1𝛽

conc

entr

atio

n(p

g/m

l)

⁎⁎⁎⁎⁎⁎⁎⁎⁎⁎

⁎⁎ ⁎

⁎

(f)

Figure 2: IL-37 suppressed expression of inflammatory factors in human amniotic epithelial cells. (a, b) Dose-dependent effects of IL-37(10 ng/ml, 50 ng/ml, and 100 ng/ml) on the mRNA and protein expression levels of IL-1β, IL-6, and TNF-α were measured by quantitativereal-time PCR and ELISA. The detection range of the ELISA kit was 7.8-1000 pg/ml. The results were normalized to β-actin. (c, d) Time-dependent effects of 100 ng/ml IL-37 (0 h, 6 h, 12 h, 24 h) on the mRNA and protein expression levels of IL-1β, IL-6, and TNF-α weredetected by quantitative real-time PCR and ELISA. The detection range of the ELISA kit was 7.8-1000 pg/ml. The results were normalizedto β-actin. (e) Transfection efficiency of IL-37 siRNA was detected by quantitative real-time PCR and western blot. The results werenormalized to β-actin. (f) The concentrations of IL-1β, IL-6, and TNF-α were measured by ELISA in the cell supernatant of the LPS-treated WISH cells after transfection with IL-37 siRNA for 48 hours and treatment with LPS and rhIL-37 for 24 hours, respectively.The detection range of the ELISA kit was 7.8-1000 pg/ml. ns: nonsignificance; ∗p < 0:05 vs. LPS; ∗∗p < 0:01 vs. LPS; ∗∗∗p < 0:001 vs.LPS; ∗∗∗∗p < 0:0001; ##p < 0:01 vs. control; ####p < 0:0001.

8 Mediators of Inflammation

MMP9

MMP2

sPTBTerm

92 kD

72 kD

(a)

Opt

ical

den

sity

of M

MPs

(arb

itrar

y un

it)

MMP20

1

2

3

MMP9TermsPTB

⁎⁎

(b)

MMP9

MMP2LPS +

+ ++

rhIL-37

92 kD

72 kD– –– –

(c)

MMP9MMP20

2

4

3

1

ControlLPSrhIL-37LPS+rhIL-37

########

Rela

tive m

RNA

expr

essio

n(a

rbitr

ary

unit)

⁎⁎⁎⁎ ⁎⁎⁎⁎

(d)

LPS+rhIL-37

MMP9

MMP2

𝛽-Tubulin

Control LPS rhIL-37

(e)

0.00.2

0.40.60.81.0

##

⁎

MM

P9 p

rote

in ex

pres

sion

(fold

chan

ge)

MM

P2 p

rote

in ex

pres

sion

(fold

chan

ge)

0.0

0.2

0.4

0.6

0.8

LPS+

rhIL

-37

rhIL

-37

LPS

Cont

rol

LPS+

rhIL

-37

rhIL

-37

LPS

Cont

rol

(f)

Figure 3: Continued.

9Mediators of Inflammation

front line for inflammatory reactions. Rupture of the fetalmembrane may cause a positive feedback loop for inflamma-tion leading to preterm birth [9]. However, the strength andflexibility of the fetal membrane are primarily dominated bythe key molecules of ECM remodeling, MMP2 and MMP9[37]. A series of studies have identified that MMP2 andMMP9 could cause the reticulin proteolysis that results inECM remodeling, which promotes membrane rupture thatlastly leads to preterm birth [8]. It has been shown that IL-1β, TNF-α, and IL-6 can induce the expression of MMP2and MMP9 in amniotic fluid and fetal membranes [38, 39].According to our results, IL-37 could inhibit the productionof IL-1β, TNF-α, and IL-6 in human amniotic epithelial cells,

so we speculate that IL-37 will inhibit the MMP2 and MMP9expressions. To clarify this hypothesis, the expressions ofMMP2 and MMP9 in WISH cells were measured after treat-ment with IL-37. Surprisingly, the MMP9 expression wasfound significantly downregulated after IL-37 treatment.However, no differences were found in MMP2 expression.According to the previous studies, IL-37 indeed affectsMMP2 and MMP9 expressions in endometriosis, atheroscle-rosis, and human lung adenocarcinoma [40–42]. But whychanges only occur in MMP9. In fact, amniotic epithelialcells exclusively produce MMP9, while amniotic mesenchy-mal cells produce only MMP2. It is therefore concluded thatMMP2 and MMP9 exhibited cell-specific expression in the

si_IL-37si_NC

92 kD

72 kD

MMP9

MMP2

(g)

0

1

Opt

ical

den

sity

of M

MPs

(arb

itrar

y un

it) 2

3

si_NC

ns

MMP9MMP2

si_IL-37

⁎

(h)

MMP2 MMP90.0

0.5

1.0

1.5

2.0

ns

Rela

tive m

RNA

expr

essio

n(a

rbitr

ary

unit)

⁎

(i)

MMP9

MMP2

𝛽-Actin

si_IL-37si_NC

(j)

si_IL-37si_NC0.00

0.05

0.25

0.10

0.15

0.20

MM

P2 p

rote

in ex

pres

sion

(fold

chan

ge)

MM

P9 p

rote

in ex

pres

sion

(fold

chan

ge)

si_IL-37si_NC0.0

0.1

0.2

0.3ns ⁎

(k)

Figure 3: Effects of IL-37 on expression and activity of MMP2 and MMP9 in fetal membranes and human amniotic epithelial cells. (a) Theactivities of MMP2 and MMP9 in fetal membranes collected from term (n = 4) and sPTB (n = 4) delivery women were measured with gelatinzymogram. (b) Statistical analysis of gelatin zymogram in the result (a). (c) The activity of MMP2 and MMP9 in WISH cells followingtreatment with LPS and rhIL-37 for 24 hours was detected with gelatin zymogram. (d) The mRNA expression levels of MMP2 and MMP9after exposed to LPS and rhIL-37 for 24 hours were detected by quantitative real-time PCR analysis. The results were normalized to β-actin. (e) The protein expression levels of MMP2 and MMP9 after exposed to LPS and rhIL-37 for 24 hours were measured by westernblot. (f) Statistical analysis of western blot in the result (e). (g) The activity of MMP2 and MMP9 between the IL-37 siRNA group and thesi_NC group was detected with gelatin zymogram. (h) Statistical analysis of gelatin zymogram in results (g). (i) The mRNA expressionlevels of MMP2 and MMP9 were measured by quantitative real-time PCR analysis in each group. The results were normalized to β-actin.(j) The protein expression levels of MMP2 and MMP9 between the IL-37 siRNA group and the si_NC group were detected by westernblot. (k) Statistical analysis of western blot in the result (j). ns: nonsignificance; ∗p < 0:05 vs. LPS; ∗∗∗∗p < 0:0001; ##p < 0:01 vs. control;####p < 0:0001.

10 Mediators of Inflammation

Control

si_NC

si_IL-37

Hoechst EdU Merge

200 𝜇m200 𝜇m200 𝜇m

200 𝜇m200 𝜇m200 𝜇m

200 𝜇m200 𝜇m200 𝜇m

(a)

20

Cont

rol

si_N

C

si_IL

-37

0

40

60

80

⁎⁎⁎⁎

Perc

enta

ge o

fEd

U-p

ositi

ve ce

lls (%

)

(b)

Control

si_NC

si_IL-37

Total cells MergeDead cells

200 𝜇m200 𝜇m200 𝜇m

200 𝜇m200 𝜇m200 𝜇m

200 𝜇m200 𝜇m200 𝜇m

(c)

Cont

rol

si_N

C

si_IL

-37

4

6

2

0

Dea

th ce

lls(d

ead/

tota

l cel

ls %

)

8 ⁎⁎⁎⁎

(d)

Control si_NC si_IL-37

107106105104107106105104107106105104

104

105 105 105

106

104

106

104

106

Annexin V FITC-AAnnexin V FITC-AAnnexin V FITC-A

PIPC

5.5-

A

Q1-UL (1.14%)

Q1-LL (94.02%) Q1-LL (93.94%)

Q1-UL (4.70%) Q1-UL (1.62%) Q1-UL (4.13%)

Q1-LR (0.14%) Q1-LR (0.13%) Q1-LR (1.74%)Q1-LL (85.93%)

Q1-UL (2.06%) Q1-UL (10.27%)

(e)

Cont

rol

si_N

C

si_IL

-37

0

5

10

15

Tota

l apo

ptot

ic ce

lls (%

)

⁎

(f)

Bcl-2

Bax

GAPDH

Control si_NC si_IL-37

(g)

Bcl-2

to G

APD

H ra

tion

(arb

itrar

y un

it)2.0

1.5

1.0

0.0

0.5

Bax

to G

APD

H ra

tion

(arb

itrar

y un

it)

0.0

0.5

1.0

1.5

Cont

rol

si_N

C

si_IL

-37

Cont

rol

si_N

C

si_IL

-37

⁎⁎

⁎⁎⁎

(h)

p-STAT3

STAT3

𝛽-Actin

si_NC si_IL-37

(i)

si_N

C

si_IL

-37

1.5

1.0

0.0

0.5

p-ST

AT3

to S

TAT3

ratio

n (a

rbitr

ary

unit) ⁎⁎

(j)

Figure 4: Effects of IL-37 on apoptosis and proliferation in human amniotic epithelial cells. (a) Cell proliferation between the IL-37 siRNAgroup and the si_NC group was measured by 5-ethynyl-20-deoxyuridine (EdU) incorporation assay. The blue color indicates the nuclei, andthe red color represents the EdU-positive nuclei. Scale bar: 200 μm. (b) Statistical analysis of EdU staining in the result (a). (c) The Live/Deathcell kit was used to analyze the number of dead cells between the IL-37 siRNA group and the si_NC group. The blue color indicates the nuclei,and the green color represents the death cells. Scale bar: 200 μm. (d) Statistical analysis of death cell rate in the result (c). (e) Cell apoptosisbetween the IL-37 siRNA group and the si_NC group was assessed by the flow cytometer. Representative pictures of flow cytometry forapoptosis rate. (f) Statistical analysis of flow cytometry in the result (e). (g) The apoptosis-related protein biomarkers were measured bywestern blot in each group. (h) Statistical analysis of western blot in the result (g). (i) The STAT3 signaling was measured by westernblot between the si_NC group and the IL-37 siRNA group. (j) Statistical analysis of western blot in the result (i). ∗p < 0:05 vs. si_NC;∗∗p < 0:01; ∗∗∗p < 0:001; ∗∗∗∗p < 0:0001.

11Mediators of Inflammation

human fetal membranes [43]. Therefore, it is understandablethat IL-37 can only regulate MMP9 instead of MMP2 inamniotic epithelial cells.

On the other hand, preterm birth is mediated by cell apo-ptosis, which may be affected by factors such as infectionsand external stimuli [9]. The risk signals stimulate Toll-likereceptors which lead to the activation of NF-κB and subse-quently increase expression of the major proinflammatoryfactors including IL-1β, IL-6, and TNF-α in fetal membranes[11]. Among them, TNF-α could band to its receptor that ini-tiates signal transduction through the TNFR-associateddeath domain (TRADD), which activates caspase activityeventually resulting in apoptosis [28]. IL-1β and IL-6 couldincrease the fragmentation of nuclei leading to apoptosis aswell [44]. It is already verified that IL-37 could limit excessiveinflammation in this study and several published articleshave revealed that IL-37 could influence apoptosis in cervicalcancer and renal carcinoma [45, 46]. Therefore, we want toinvestigate whether IL-37 can inhibit the apoptosis of amni-otic epithelial cells in preterm birth. As is known to all, Baxand Bcl-2 are homologous proteins in the mitochondrialapoptosis pathway. The imbalance of Bax/Bcl-2 is capableof leading to apoptosis. Fortunato et al. have reported thatthe expression of Bax was increased, while Bcl-2 wasdecreased in the fetal membranes of the sPTB pregnancies[47]. A recent study suggested that IL-37 could inhibit Bcl-2 through STAT3 signaling [48]. Hence, we speculate that

IL-37 is also able to suppress apoptosis by affecting thebalance of Bax/Bcl-2 via the STAT3 pathway. According toour study, the results showed that the downregulation ofIL-37 could promote apoptosis by increasing the expressionof proapoptotic protein Bax and decreasing antiapoptoticprotein Bcl-2 in WISH cells. The consistent results wereshown in flow cytometry analysis. Furthermore, as shownin Figure 4(i) and Supplementary Figure 1, bothknockdown and treatment with IL-37 both impactedSTAT3 signaling. Thus, it is proved that IL-37 can affectpreterm birth by regulating apoptosis through the STAT3signaling pathway.

According to the above results, IL-37-mediated anti-inflammatory effects play a critical role in sPTB. Next, wetentatively explore its specific molecular mechanism. LPS,the component of Gram-negative bacteria, can generally acti-vate the NF-κB signaling pathway, which provokes a widevariety of inflammation responses and triggers the produc-tion of proinflammatory cytokines, including IL-1β, IL-6,and TNF-α [49]. According to our results (Figure 5(a)), theNF-κB signaling pathway was activated after stimulated byLPS. Meanwhile, as shown in Figure 2(f), IL-1β, IL-6, andTNF-α were all increased, particularly IL-6. Previous studieshave indicated that IL-6, the most well-known traditionalactivator of STAT3, was a hallmark of many human chronicinflammatory states, including sepsis, rheumatoid arthritis(RA), and inflammatory bowel disease (IBD) [50]. Hence, it

p-NF-𝜅B p65

NF-𝜅B p65

p-STAT3

STAT3

𝛽-Actin

rhIL-37LPS + +

+–

– –

(a)

2.0

1.5

0.5

1.0

0.0

1.5

0.5

1.0

LPSControl

####

0.0LPS+

rhIL-37LPSControl LPS+

rhIL-37

p-ST

AT3

to S

TAT3

ratio

n(a

rbitr

ary

unit)

p-N

F-𝜅

B p6

5 to

NF-𝜅

B p6

5 ra

tion

(arb

itrar

y un

it)

⁎⁎

(b)

p-STAT3

STAT3

𝛽-Actin

rhIL-37rhIL-6 + +

+–

– –

(c)

rhIL-6Control

1.5

1.0

0.0

0.5

##

rhIL-6+rhIL-37

p-ST

AT3

to S

TAT3

ratio

n(a

rbitr

ary

unit)

⁎⁎

(d)

Figure 5: IL-37 exerts its biological functions through the NF-κB and IL-6/STAT3 signaling pathway. (a) The phosphorylation of NF-κB,STAT3, and total amounts of NF-κB, STAT3 were detected by western blot after administrated by LPS and LPS+rhIL-37 for 12 hours. (b)Densitometry quantifications of p-NF-κB to T-NF-κB and p-STAT3 to T-STAT3 level were shown. (c) Western blot analysis of p-STAT3and T-STAT3 expression was presented following treatment with rhIL-6 and rhIL-37+rhIL-6 for 12 hours. (d) Densitometryquantification of p-STAT3 to T-STAT3 level was demonstrated. ∗p < 0:05 vs. LPS; ∗∗p < 0:01 vs. rhIL-6; ##p < 0:01 vs. control.

12 Mediators of Inflammation

could conceivably be hypothesized that LPS can exert itseffects through the IL-6/STAT3 signaling pathway indirectlyin sPTB. Basing on our results, STAT3 was upregulated aftertreated with IL-6. Therefore, it is understandable that LPScould activate phosphorylation of the STAT3 pathwaythrough IL-6 indirectly. Previous studies suggested that IL-37 could inhibit the activation of the NF-κB and IL-6/STAT3signaling pathway and suppressed the occurrence of theinflammatory response in Myasthenia Gravis (MG) by affect-ing Tfh and B cells, and in lung cancer by affecting A549 cells[46, 51]. However, the anti-inflammatory effect of IL-37 viathe NF-κB and IL-6/STAT3 pathway in preterm birthremains elusive. The results showed that the expression ofp-NF-κB and p-STAT3 was significantly increased in theLPS-treated group, whereas significantly decreased afterincubation with IL-37. Thus, it can be suggested that LPSactivates the NF-κB signaling pathway and then to triggerthe production of proinflammatory factors (mainly IL-6),which subsequently phosphorylates STAT3. However, theabove effects are reversed by IL-37.This study was confirmedby a previous study, further stating that IL-37 harnesses theanti-inflammatory properties of the signaling molecule NF-κB and STAT3 [52].

Although we have conducted in vitro studies to verify ourhypothesis, there are still some details that need furtherrefinement. For the consistency of the results, the amnioticepithelial cells applied in our in vitro study are a cell lineinstead of the primary amniotic epithelial cells. Thus, we planto address some of these issues in the next experiment, andin vivo study is also being scheduled. Furthermore, Master-son et al. have indicated that mesenchymal stem cells (MSCs)and their derivatives are incapable of modulating IL-37, andMSCs expressing IL-37 may have an enhanced therapeuticefficacy [53]. Actually, the fetal membrane, especiallyamnion, is full of MSCs, which may produce efficient IL-37to maintain the immune balance at the maternal-fetal inter-face. Hence, IL-37 may be a therapeutic target for sPTB inthe near future. However, more researches are needed.

5. Conclusion

In summary, the anti-inflammatory effect of IL-37 in fetalmembranes by suppressing the NF-κB and IL-6/STAT3 sig-naling pathway can inhibit apoptosis and remodeling ofECM and participate in preterm birth. These findings mayfurther enrich the theoretical strategies for preterm birth.

Data Availability

The data used to support the findings of this study are avail-able from the corresponding author upon request.

Disclosure

Lulu Wang and Zheng Liu should be considered joint firstauthors.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Acknowledgments

The following individuals are gratefully acknowledged fortheir assistance: E Gong, Lian Yu, Juan Qin, and ChuanyueFu. The authors are also grateful to all the pregnant womenwho participated in our study and the support of the FirstAffiliated Hospital of ChongqingMedical University, Chong-qing, China. Furthermore, we would like to appreciate thesupport from “111 program” of the Ministry of Educationof PRC and the State Administration of Foreign ExpertsAffairs of PRC and the cooperation of the Obstetrics, Nurseand Midwifery staff of the First Affiliated Hospital of Chong-qing Medical University, Chongqing, China. This work wassupported by grants from the National Key R&D Programof China (No. 2016YFC1000407), the Key Research Programof Chongqing Association of Science and Technology (No.cstc2017shms-zdyfX0034), and the National Natural ScienceFoundation of China (No. 81801483).

Supplementary Materials

Figure S1: effects of IL-37 on apoptosis through IL-6 inhuman amniotic epithelial cells. (a) The expression of Baxand Bcl-2 was detected by western blot after treatment withrhIL-6 and rhIL-6+rhIL-37 for 12 hours. (b) Statistical anal-ysis of western blot in the result (a). ∗∗p < 0:01 vs. rhIL-6;#p < 0:05 vs. control; ###p < 0:001. (Supplementary Materials)

References

[1] S. Chawanpaiboon, J. P. Vogel, A.-B. Moller et al., “Global,regional, and national estimates of levels of preterm birth in2014: a systematic review and modelling analysis,” The LancetGlobal Health, vol. 7, no. 1, pp. e37–e46, 2019.

[2] H. A. Frey and M. A. Klebanoff, “The epidemiology, etiology,and costs of preterm birth,” Seminars in Fetal and NeonatalMedicine, vol. 21, no. 2, pp. 68–73, 2016.

[3] F. Bodeau-Livinec, N. Marlow, P.-Y. Ancel, J. J. Kurinczuk,K. Costeloe, and M. Kaminski, “Impact of intensive care prac-tices on short-term and long-term outcomes for extremelypreterm infants: comparison between the British Isles andFrance,” PEDIATRICS, vol. 122, no. 5, pp. e1014–e1021, 2008.

[4] F. Wu, G. Liu, Z. Feng et al., “Short-term outcomes ofextremely preterm infants at discharge: a multicenter studyfrom Guangdong Province during 2008–2017,” BMC Pediat-rics, vol. 19, no. 1, p. 405, 2019.

[5] R. Romero, S. K. Dey, and S. J. Fisher, “Preterm labor: one syn-drome, many causes,” Science, vol. 345, no. 6198, pp. 760–765,2014.

[6] N. Gomez-Lopez, D. StLouis, M. A. Lehr, E. N. Sanchez-Rodriguez, and M. Arenas-Hernandez, “Immune cells in termand preterm labor,” Cellular &Molecular Immunology, vol. 11,no. 6, pp. 571–581, 2014.

[7] A. K. Boyle, S. F. Rinaldi, J. E. Norman, and S. J. Stock,“Preterm birth: inflammation, fetal injury and treatmentstrategies,” Journal of Reproductive Immunology, vol. 119,pp. 62–66, 2017.

13Mediators of Inflammation

[8] N. Yin, H. Wang, H. Zhang et al., “IL-27 induces a pro-inflammatory response in human fetal membranes mediatingpreterm birth,” International Immunopharmacology, vol. 50,pp. 361–369, 2017.

[9] R. Menon and S. J. Fortunato, “Infection and the role ofinflammation in preterm premature rupture of the mem-branes,” Best Practice & Research Clinical Obstetrics & Gynae-cology, vol. 21, no. 3, pp. 467–478, 2007.

[10] R. Menon, “Human fetal membranes at term: dead tissue orsignalers of parturition?,” Placenta, vol. 44, pp. 1–5, 2016.

[11] M. Phillippe, “Cell-free fetal DNA, telomeres, and the sponta-neous onset of parturition,” Reproductive Sciences, vol. 22,no. 10, pp. 1186–1201, 2015.

[12] C. Gillaux, C. Méhats, D. Vaiman, D. Cabrol, andM. Breuiller-Fouché, “Functional screening of TLRs in human amnioticepithelial cells,” The Journal of Immunology, vol. 187, no. 5,pp. 2766–2774, 2011.

[13] Y. Abe, M. Komatsubara, M. Saito et al., “Activin A is stimu-lated by tumor necrosis factor-alpha and modulates collagengene expression in human amniotic cells,” Journal of Endocri-nological Investigation, vol. 36, no. 7, pp. 515–520, 2013.

[14] B. Huang, A. N. Faucette, M. D. Pawlitz et al., “Interleukin-33-induced expression of PIBF1 by decidual B cells protectsagainst preterm labor,” Nature Medicine, vol. 23, no. 1,pp. 128–135, 2017.

[15] S. Dambaeva, S. Schneiderman, M. K. Jaiswal et al., “Interleu-kin 22 Prevents lipopolysaccharide- induced Preterm Labor inmice†,” Biology of Reprodictive, vol. 98, no. 3, pp. 299–308,2018.

[16] G. Cavalli and C. A. Dinarello, “Suppression of inflammationand acquired immunity by IL-37,” Immunological Reviews,vol. 281, no. 1, pp. 179–190, 2018.

[17] C. A. Dinarello, C. Nold-Petry, M. Nold et al., “Suppression ofinnate inflammation and immunity by interleukin-37,” Euro-pean Journal of Immunology, vol. 46, no. 5, pp. 1067–1081,2016.

[18] H. Jia, J. Liu, and B. Han, “Reviews of interleukin-37: func-tions, receptors, and roles in diseases,” BioMed Research Inter-national, vol. 2018, 14 pages, 2018.

[19] A. M. Ellisdon, C. A. Nold-Petry, L. D’Andrea et al., “Homodi-merization attenuates the anti-inflammatory activity of interleu-kin-37,” Science Immunology, vol. 2, no. 8, p. eaaj1548, 2017.

[20] L. Ye, L. Ji, Z. Wen et al., “IL-37 inhibits the production ofinflammatory cytokines in peripheral blood mononuclear cellsof patients with systemic lupus erythematosus: its correlationwith disease activity,” Journal of Translational Medicine,vol. 12, no. 1, p. 69, 2014.

[21] J. Southcombe, I. Sargent, C. Redman, and I. Granne, “IL-1family cytokines and their regulatory proteins in normal preg-nancy and pre-eclampsiaIL-1 family cytokines and their inhib-itory proteins in normal pregnancy and pre-eclampsia,”Journal of Reproductive Immunology, vol. 101-102, pp. 22-23,2014.

[22] Z. Yu, J. Liu, R. Zhang et al., “IL-37 and 38 signalling in gesta-tional diabetes,” Journal of Reproductive Immunology, vol. 124,pp. 8–14, 2017.

[23] J. Jiang, K. Yu, Z. Jiang, and M. Xue, “IL-37 affects the occur-rence and development of endometriosis by regulating the bio-logical behavior of endometrial stromal cells through multiplesignaling pathways,” Biological Chemistry, vol. 399, no. 11,pp. 1325–1337, 2018.

[24] R. L. Goldenberg, J. F. Culhane, J. D. Iams, and R. Romero,“Preterm Birth 1: Epidemiology and causes of preterm birth,”Lancet, vol. 29, no. 1, pp. 6-7, 2009.

[25] P. Chaemsaithong, R. Romero, N. Docheva et al., “Compari-son of rapid MMP-8 and interleukin-6 point-of-care tests toidentify intra-amniotic inflammation/infection and impend-ing preterm delivery in patients with preterm labor and intactmembranes,” The Journal of Maternal-Fetal & Neonatal Med-icine, vol. 31, no. 2, pp. 228–244, 2017.

[26] Y. Tanaka, H. Narahara, N. Takai, J. Yoshimatsu, T. Anai, andI. Miyakawa, “Interleukin-1β and interleukin-8 in cervicovagi-nal fluid during pregnancy,” American Journal of Obstetricsand Gynecology, vol. 179, no. 3, pp. 644–649, 1998.

[27] M. Pandey, M. Chauhan, and S. Awasthi, “Interplay of cyto-kines in preterm birth,” Indian Journal of Medical Research,vol. 146, no. 3, pp. 316–327, 2017.

[28] S. J. Fortunato, R. Menon, and S. J. Lombardi, “Role of tumornecrosis factor-α in the premature rupture of membranes andpreterm labor pathways,” American Journal of Obstetrics andGynecology, vol. 187, no. 5, pp. 1159–1162, 2002.

[29] T. Weissenbacher, R. P. Laubender, S. S. Witkin et al., “Diag-nostic biomarkers of pro-inflammatory immune-mediatedpreterm birth,” Archives of Gynecology and Obstetrics,vol. 287, no. 4, pp. 673–685, 2013.

[30] Y. Liu, Y. Liu, R. Zhang, L. Zhu, and Z. Feng, “Early- or mid-trimester amniocentesis biomarkers for predicting pretermdelivery: a meta-analysis,” Annals of Medicine, vol. 49, no. 1,pp. 1–10, 2016.

[31] M. Nadeau-Vallée, D. Obari, C. Quiniou et al., “A critical roleof interleukin-1 in preterm labor,” Cytokine & Growth FactorReviews, vol. 28, pp. 37–51, 2016.

[32] C. E. Gallenga, F. Pandolfi, Caraffa al et al., “Interleukin-1 fam-ily cytokines and mast cells activation and inhibition,” Journalof Biological Regulators and Homeostatic Agents, vol. 33, no. 1,pp. 1–6, 2019.

[33] L. Franza, V. Carusi, S. Altamura et al., “Interrelationshipbetween inflammatory cytokines (IL-1, IL-6, IL-33, IL-37)and acquired immunity,” Journal of Biological Regulators andHomeostatic Agents, vol. 33, no. 5, pp. 1321–1326, 2019.

[34] M. F. Nold, C. A. Nold-Petry, J. A. Zepp, B. E. Palmer,P. Bufler, and C. A. Dinarello, “IL-37 is a fundamental inhibi-tor of innate immunity,” Nature Immunology, vol. 11, no. 11,pp. 1014–1022, 2010.

[35] A. Wakabayashi, K. Sawada, M. Nakayama et al., “Targetinginterleukin-6 receptor inhibits preterm delivery induced byinflammation,” MHR: Basic science of reproductive medicine,vol. 19, no. 11, pp. 718–726, 2013.

[36] A. Toda, K. Sawada, T. Fujikawa et al., “Targeting inhibitor ofκB kinase β prevents inflammation-induced preterm deliveryby inhibiting IL-6 production from amniotic cells,” The Amer-ican Journal of Pathology, vol. 186, no. 3, pp. 616–629, 2016.

[37] J. V. Cockle, N. Gopichandran, J. J. Walker, M. I. Levene, andN. M. Orsi, “Matrix metalloproteinases and their tissue inhib-itors in preterm perinatal complications,” Reproductive Sci-ences, vol. 14, no. 7, pp. 629–645, 2016.

[38] F. Vadillo-Ortega, D. W. Sadowsky, G. J. Haluska et al., “Iden-tification of matrix metalloproteinase-9 in amniotic fluid andamniochorion in spontaneous labor and after experimentalintrauterine infection or interleukin-1β infusion in pregnantrhesus monkeys,” American Journal of Obstetrics and Gynecol-ogy, vol. 186, no. 1, pp. 128–138, 2002.

14 Mediators of Inflammation

[39] F. Vadillo-Ortega and G. Estrada-Gutiérrez, “Role of matrixmetalloproteinases in preterm labour,” BJOG: An Interna-tional Journal of Obstetrics & Gynaecology, vol. 112, Suppl 1,pp. 19–22, 2005.

[40] A. Ladenburger, M. Seehase, B. W. Kramer et al., “Glucocorti-coids potentiate IL-6-induced SP-B expression in H441 cellsby enhancing the JAK-STAT signaling pathway,” AmericanJournal of Physiology-Lung Cellular and Molecular Physiology,vol. 299, no. 4, pp. L578–L584, 2010.

[41] J. Liu, J. Lin, S. He et al., “Transgenic overexpression of IL-37protects against atherosclerosis and strengthens plaque stabil-ity,” Cellular Physiology and Biochemistry, vol. 45, no. 3,pp. 1034–1050, 2018.

[42] Y.-H. Chen, B.-Y. Zhou, G.-C. Wu et al., “Effects of exogenousIL-37 on the biological characteristics of human lung adeno-carcinoma A549 cells and the chemotaxis of regulatory Tcells,” Cancer Biomarkers, vol. 21, no. 3, pp. 661–673, 2018.

[43] C. Janzen, S. Sen, M. Y. Y. Lei, M. Gagliardi de Assumpcao,J. Challis, and G. Chaudhuri, “The role of epithelial to mesen-chymal transition in human amniotic membrane rupture,”The Journal of Clinical Endocrinology & Metabolism,vol. 102, pp. 1261–1269, 2016.

[44] S. J. Fortunato and R. Menon, “IL-1β is a better inducer of apo-ptosis in human fetal membranes than IL-6,” Placenta, vol. 24,no. 10, pp. 922–928, 2003.

[45] S. Wang, W. An, Y. Yao et al., “Interleukin 37 expressioninhibits STAT3 to suppress the proliferation and invasion ofhuman cervical cancer cells,” Journal of Cancer, vol. 6,no. 10, pp. 962–969, 2015.

[46] M. Jiang, Y. Wang, H. Zhang et al., “IL-37 inhibits invasionand metastasis in non-small cell lung cancer by suppressingthe IL-6/STAT3 signaling pathway,” Thoracic Cancer, vol. 9,no. 5, pp. 621–629, 2018.

[47] S. J. Fortunato, R. Menon, C. Bryant, and S. J. Lombardi, “Pro-grammed cell death (apoptosis) as a possible pathway tometalloproteinase activation and fetal membrane degradationin premature rupture of membranes,” American Journal ofObstetrics and Gynecology, vol. 182, no. 6, pp. 1468–1476,2000.

[48] Y. Jiang, Y. Wang, L. Liang et al., “IL-37 mediates the antitu-mor activity in renal cell carcinoma,” Medical Oncology,vol. 32, no. 11, 2015.

[49] S. Akira and K. Takeda, “Toll-like receptor signalling,” NatureReviews Immunology, vol. 4, no. 7, pp. 499–511, 2004.

[50] C. J. Greenhill, S. Rose-John, R. Lissilaa et al., “IL-6 trans-signaling modulates TLR4-dependent inflammatory responsesvia STAT3,” The Journal of Immunology, vol. 186, no. 2,pp. 1199–1208, 2011.

[51] Z. Liu, L. Zhu, Z. Lu et al., “IL-37 represses the autoimmunityin myasthenia gravis via directly targeting follicular Th and Bcells,” The Journal of Immunology, vol. 204, no. 7, pp. 1736–1745, 2020.

[52] Z. Zhang, J. Zhang, P. He, J. Han, and C. Sun, “Interleukin-37suppresses hepatocellular carcinoma growth through inhibit-ing M2 polarization of tumor-associated macrophages,”Molecular Immunology, vol. 122, pp. 13–20, 2020.

[53] C. H. Masterson, G. F. Curley, and J. G. Laffey, “Modulatingthe distribution and fate of exogenously delivered MSCs toenhance therapeutic potential: knowns and unknowns,” Inten-sive Care Medicine Experimental, vol. 7, no. S1, p. 41, 2019.

15Mediators of Inflammation